286 REV. PROE. G. HENSLOW ON THE [Aug. I907, 



derived from response to the influence of water ; as in the arrest of 

 the primary root, disrupted vascular bundles of the stem, phyllo- 

 dineous, parallel-veined leaves, etc. 



To return to the Equisetales. With regard to foliage, fig. 28 (op. 

 cit. p. 65), of a i branch bearing whorled leaves ' of ArcJiceocalamites 

 radiatus, is extraordinarily suggestive of existing submerged dicoty- 

 ledonous leaves. It is precisely what one would expect if a leaf, for 

 example, of a Sphenophyllum grew under water ; such degeneration 

 is the result of response to water. Since, however, this genus is 

 said to be arborescent, although to what height it grew I do not 

 know, the aquatic type of dissected leaf presumably has been re- 

 tained, that is, on the assumption that its ancestor was a submerged 

 herb ; just as an aquatic species of Ranunculus retains the hereditary 

 dissected, submerged form of leaf, whenever it grows on land and in 

 air. Palm-trees have, similarly, retained features in their foliage 

 characteristic of an aquatic origin. Indeed, several species of Palm, 

 as stated, form tropical swamp-forests. It may be added that a 

 branched midrib and veins of Sphenophyllum demand a previous 

 complete leaf with the interstices filled up, because the 4 object ' is 

 to offer a broad expanse of blade to the sun ; consequently there is 

 no doubt of its being descended from some such leaf as that of, say, 

 a Cingularia. 



With regard to the xerophytic character of the leaf-rudiments of 

 Catamites and Equisetum, I would suggest the following hypothetical 

 evolution : — The first stage was that of a Cingularia, or Spheno- 

 phyllum, or other type of a complete blade, being a terrestrial form. 

 Becoming aquatic, this was reduced to dichotomous ribs and veins 

 only, as in ArcJiceocalamites. Then, becoming exposed to very xero- 

 phytic conditions, this was reduced to the petioles only, thus forming 

 the ' points ' on the whorled sheaths. 



Modern types are seen in Casuarina, some Eestiaceous plants, and 

 Polygonum equisetifolium. The genera of Restiaceai abound in the 

 ' Flats ' near Cape Town. They occur like rushes in the swamp 

 during the wet season, but are perfect xerophytes in the dry season. 

 They thus illustrate Dr. Schimper's ' tropophytes/ 



Turning next to Lycopodiales, figs. 51 & 53 (op. cit. pp. 126 & 133) 

 of Lepidodendron exhibit a compact xylem and cortex, without any 

 lacunaa. Pig. 56 (p. 141) is also a radial section of compact xylem 

 with no large vessels, and none are disorganized. 



A tall tree requires much rigidity in its stem, and these illus- 

 trations show clearly how Lepidodendra were well qualified for 

 supporting the weight of their gigantic trunks. But, when trees 

 are hygrophytes, as species of willows and poplars, the wood is in a 

 much reduced state, comparatively speaking, such trees having 

 numerous large vessels scattered indiscriminately through the wood- 

 fibres, these too being correspondingly enfeebled. Of course, 

 nothing at all resembling poplar-wood occurs in the Carboniferous 

 trees, so far as the figures in Dr. Scott's book testify ; but one would 



